August 20th, 2016

Signal Tracing and Injection Basics

Disclaimer
The content of this website and blog has been created and posted here solely for the purposes of reading interest by Rob's Radio-Active, LLC. The content is the thoughts of the author and is not intended to be an authoritative guide. There may be errors and omissions. Electronic and electrical devices pose a potential hazard to life and property if improperly handled, used, or serviced. The author cannot possibly think of, and warn of, all the ways one can get into trouble with electricity, and recommends one learn all of the technical and safety aspects of electrical/electronics before attempting to work with or use any of it. There are hazardous voltages and currents involved in almost all electrical and electronic equipment—especially vacuum tube devices or anything that plugs into mains current for operation. Furthermore, vintage equipment was never designed or built to be up the safety standards of modern times.
The purpose of signal tracing, or another form of it, signal injection, is to locate the defective stage of a malfunctioning electronic system such as a radio or audio amplifier/guitar amplifier as efficiently as possible. Once the defective stage is determined, methods of checking individual voltages and components within that stage can be employed to isolate the failure to an individual component or wiring connection for repair. Since one must be working with a live circuit, extra caution must be employed to avoid electrical shock or fire, or damage to the equipment. Be sure you know what you are doing, and are competent in general electrical methods and safety. This is especially true when working on vintage AC/DC radio receivers that use a "hot" chassis. The chassis itself may be at power line voltage depending on the power cord polarity, and an isolation transformer is a must when working on those. It is also good practice to work with one hand to avoid making a current path across your body, and to make or remove test connections with the unit unplugged from the power. Beware that in a malfunctioning unit, voltages may be present where they are not supposed to be, and also that the capacitors may retain a dangerous charge for a long time after power is removed from the unit being serviced. These are just a few of the precautions that must be taken.

In this example I will be using a simple five-tube AC/DC type AM radio receiver, but the principles remain the same even on a solid state AM/FM stereo or an audio amplifier as long as you can find the test connection points. It is usually necessary, in any case, to have and be able to read the schematic and know how to find the test points. At the very least, a tube manual or datasheet for the tube in the device under test should be consulted for the pin numbers that connect to the various elements. For example: tube pins numbers are counted clockwise from the bottom starting from the right side of the gap in the pins for the common 7 and 9 pin miniature tubes and from the locating key on the center stud of octal or loctal type 8-pin tubes.

Generally speaking, when signal tracing, one starts at the front end—the antenna end of a receiver or the input end of an amplifier—by injecting a signal from a signal generator and test for signal presence, strength, and quality working toward the speaker end. When signal injection testing using only a signal generator, it is best to inject an appropriate test signal at different points, starting at the speaker end, and monitoring the output on the receiver or amplifier's own speaker or output to determine at what stage the signal becomes lost or degraded. We will start with tracing with a signal tracer, my particular model being an EICO model 147. The fastest way to track down a malfunctioning stage is actually not this way—at least not for the experienced troubleshooter—the fastest way is to start in the electronic middle (such as the volume control) and then further divide each remaining half in half again. For our purposes, let's pretend you do not work at a hypothetical vintage radio and amplifier repair shop where you have to fix ten sets a day to keep your job. Let's keep it simple and maybe learn something along the way, and start at the ends of the device under test. 

One advantage of using a tracer for a radio receiver, is that it is possible to use a good, strong radio station instead of injecting our own signal at the antenna end, but if the receiver is inoperative, it can be difficult to find a station at all, and you must find one if possible with the tracer placed at some point in the receiver—such as the convertor output—where there is sufficient amplification to operate the tracer. In the case of an audio amplifier, any normal signal source can be used at the input, but it is usually best to inject an audio signal with a frequency adjustable audio signal generator if one is available. An audio signal generator, unlike most radio signal generators, will normally have a wide variety of frequencies that can be injected into an audio amplifier to test its response over the entire audio range of 20Hz to 20kHz. Further, a dedicated audio signal generator usually puts out a nice clean sine (or square) wave which makes it easier to detect moderate to severe harmonic distortion by ear or by oscilloscope.

The general purpose radio signal generator that we use on a radio puts out a choice of modulated or unmodulated radio frequency voltage or a single audio tone of 400Hz or so. The signal quality is usually rich in harmonics by design on the RF so that it can be used to inject signals for FM or shortwave receivers that require a frequency outside of the range of the basic generator fundamental frequencies at a lower harmonic, say for example 50MHz generator setting, and still pick up a strong response in the receiver at 100Mhz and 150MHz, and so on. The one audio tone that is usually available is normally a decent sine wave, but not entirely clean either. Using a radio signal generator to troubleshoot a receiver by signal injection can show up severe distortion in a defective receiver stage, at that test frequency, but leaves much to be desired compared to listening into a strong radio station with a signal tracer with reasonably good audio fidelity when checking the sound quality passing through the various stages within the receiver.

Using both a signal generator and tracer together is often desirable or required to detect a signal in the "front end" of a receiver—the RF amplifier (if present) and convertor stage— due to the low signal levels there. The tracer, like any test instrument, will load the circuit down to some extent because it draws a small amount of power from the circuit under test. A generator can be used to inject a modulated RF signal into the antenna circuit with about ten turns of insulated wire, about eight to ten inches in diameter  and placed about a foot from the antenna coil. With an appropriately strong generator output setting, most tracers will have no trouble finding the tone at the RF amp/convertor. You can even inject around 0.1 volt RMS radio frequency by direct contact into the antenna circuit if necessary to operate the tracer, but must usually be turned down or removed for tracing further into the receiver to avoid overloading the amplifiers. Also, a signal tracer (oscilloscope, or any other test instrument) also has a certain amount of capacitance that will tend to detune or pull a receiver's tuned circuits off frequency a little—or sometimes a lot. For relative measurements, however, usually rocking the tuning dial a little for maximum tracer output is sufficient. In passing, for more quantitative voltage measurements using an oscilloscope the tuner and oscillator or IF circuits should be tweaked for maximum output with the scope in place to balance out the added capacitance of the scope and give an accurate voltage indication. They must be put back into adjustment after the scope probe is removed. For accurate frequency measurements using a scope, say the oscillator frequency, the scope probe must be loosely coupled to the circuit by placing the probe in proximity to the circuit but not making direct contact. Putting  a coil of several turns of insulated wire across the probe's tip and ground lead, and moving this close to the circuit under test, will increase sensitivity. Obviously, the RF voltage measurements of that circuit will have to be done separately as before. Just some hints.

A signal tracer used to be on just about every service bench in the electronics service industry, and I am a fan of them. Although an oscilloscope or even a meter with an radio frequency demodulator probe can also be used to follow a signal through a device under test, I find the tracer to be a fastest way to find and monitor a signal through the circuitry and easily shows distortion from a partially malfunctioning stage. A "homebrew " tracer is perfectly usable as well, and can be fashioned from a small audio amplifier and a demodulator probe for use in the radio frequency sections of a receiver. The demodulator or "RF probe" in its  simplest form is a probe with a diode and a capacitor arranged to turn the radio frequency into an audio frequency that the audio amplifier can use. I prefer to use a vintage factory made unit because of the several extra features, such as the ability to substitute its speaker and/or output transformer for a suspect unit, a built in wattmeter, etc. See figure 1.

Figure 1 Signal Tracer

This particular model uses an "eye tube" visible in the upper right hand corner as a relative signal strength indicator. Some signal tracers use a meter movement, and many have neither. The speaker loudness at a particular gain control setting giving the indication of relative signal strength. Neither the eye tube nor the meter are particularly important to have, in my opinion. Essentially, the common signal tracer design such as this EICO 147 is just a high gain audio amplifier with an input for audio signals through a pair of straight wire probes. For "listening in" on the radio frequency signals, there is provided a demodulator probe with a shielded cable that connects to the RF input. This RF probe simply rectifies and filters the amplitude modulated radio frequency signals into an audio signal that goes directly to the same audio amplifier as used with direct audio signals. Interestingly, and sometimes useful, there is also a "noise test" function on this model that puts out a relatively high (about 100V) DC voltage through an internal current limiting resistor. When switched to the noise test mode, this DC voltage is applied to the audio input jacks  and probes. The probes are typically used to apply the high voltage across capacitors, resistors, or suspected bad solder connections. If the DC current flows smoothly (or not at all in the case of a good, non-leaking capacitor) the tracer's amplifier remains silent. If there is a rough or make and break current flow through a component or connection suspected of causing noisy operation, the audio amplifier can pick this up and make a grating, popping, or static noise. 

In the following example of signal tracing through a basic vintage AM radio receiver, The key connection points are numbered in the order that I prefer to test them. Since it is usually best to use to connect the tracer probes to the safer, low voltage (usually) grids of the tubes rather than to the high voltage plates, there are a couple of places where there is back tracking if and when needed. See figures 2 through 4 below.

Figure 2 Test Points on Schematic Diagram

Figure 3 Test Points

Figure 4 Test Points Continued

Assuming the use of a signal generator rather than a radio station, Point 1 is where a radio frequency signal would be applied. This is the point where the antenna loop feeds the signal to the convertor (mixer). Since a low level signal must be applied here to avoid overloading the receiver, it is often best to loop a five or ten turns of insulated wire to the output of the generator and place this about a foot away from the antenna loop. If the generator has an attenuator that goes nearly to zero output, it can be connected directly to point 1 through a .001 microfarad capacitor. The capacitor helps match the impedance of the antenna loop to the generator. Often times a 270 to 390 ohm resistor in series with the connection works as well.

An advantage of using a signal generator, if available, is that you can choose to apply either a broadcast band frequency or an IF frequency (usually 455kHz) at point 1. This can tell you quickly whether or not the receiver's local oscillator is running. For example, If putting an AM modulated signal into point 1 that is within the RF range—550 kHz to 1600 KHz for an AM receiver—yields nothing from the receiver's own speaker, but putting in a 455 kHz IF frequency signal passes a tone through the speaker, it is probably that the local oscillator is not running—a fairly common failure. If this is the case then we know the IF amplifier and audio amplifier are working. The oscillator needs to be running to mix with the 550 kHz to 1600 kHz antenna signal to produce the 455 kHz IF signal that the IF amplifier will respond to. By injecting directly a modulated IF signal of sufficient strength at the antenna connection point, enough will be forced through the tuning circuit and pass onto the IF amplifier and the rest of the radio circuit. A further check on the oscillator function is to check the oscillator grid (pin 1) of the 12BE6 convertor tube for a negative voltage relative to ground of several volts. This voltage is the bias voltage for the tube that is created by the oscillator signal drawing a small amount of current through the convertor grid. If the oscillator is not running, the voltage will be very low or not there at all. For this measurement, a high impedance meter such as a VTVM or a digital meter with a 1 megohm resistor at the tip of the test probe should be used to avoid loading out any bias voltage which could kill the oscillator if it was, in fact, running before. Often what happens is that the plates of the oscillator section of the tuning capacitor become shorted together, at least at some tuning positions, either due to bent plates touching or conductive debris between the plates. It is important to check the oscillator function at several points across the tuning dial to determine if the oscillator comes and goes. An weak or intermittent oscillator will also tend to run at the lower frequency dial settings and die out at the high end. Other causes of oscillator failure are a weak convertor tube, shorted or open oscillator coils (not very common) or any other component in the circuit being out of specification. Any resistors can open or drift far out of tolerance and disturb the voltages needed to run the circuit. Capacitors can open, leak current, or short completely. But having isolated the problem in this case to the local oscillator, we need only check voltages and components in that area for now.

By the way, If your signal generator frequency dial is accurate, you can also check your oscillator's frequency with it another, more direct, way by using it together with a signal tracer. The receiver is set to 1 MHz, for example (and any tunable frequency will do) by using the generator's output to set the receiver's tuner dial accurately. Then the generator's signal output lead is clipped to the tracer's RF input lead. The two are moved close to, but not touching the oscillator coil and the generator is adjusted to output approximately the IF frequency higher than the tuning dial—1455 kHz in this example—and then finely tuned for "zero beat." Along the way there will be points where harmonics will make a squeal in the tracer's speaker that come onto the speaker, drop in frequency to nothing, then rise in frequency and go away again as the generator is tuned through them, but the loudest one of these will be at the fundamental oscillator frequency, and when the generator is at the quiet zero beat position at the center of that loudest set of squeals, its dial will be set on the frequency that the receiver's oscillator is running.

Strictly speaking, the previous example was a signal injection or signal substitution test for isolating a defective stage, and the tracer didn't need to be used at all. But let's assume nothing got through to the receiver's speaker with either at RF or IF frequency applied to point 1. You would next use the signal tracer's RF probe at point 2 and listen for signal. If you are not using a generator to make your own signal, you have to tune around and try to find a strong station. If using a generator, set the receiver's tuner and generator to the same frequency, perhaps near the center of the band or about 1 MHz, and tune back and forth to try to find the tone from the modulated generator signal. If the tone is found clearly here, then the oscillator/convertor must be working because the IF transformer would not pass any significant signal to point 2 unless it had been converted to 455 kHz—or whatever the IF frequency of the receiver is. If the signal is reasonably loud but raspy and distorted, try turning down the generator output a bit. It is probably overloading the convertor.  If there is no signal here, then you must work backward to the plate pin of the convertor tube or the input of the IF transformer, whichever is most convenient. There will be a fairly high voltage there, as on any plate connection point, so be aware of that. If the signal isn't making it through the IF transformer, it is either defective or badly mistuned. IF transformers rarely mistune themselves that far by simple aging, but it is not uncommon to have had someone in there before you twiddling with adjustment screws. Note that in a normally operating receiver there will usually be a bit of an apparent signal strength loss from the input to the output side of an IF transformer because of the loading effect of the tracer, and the fact that the primary and secondary windings of the transformer are pretty  loosely coupled together, magnetically. If, however, you place the tracer's RF probe at point 1 where the signal generator is connected, and then move the probe to point 2, after the convertor, there should normally be a little bit of an increase in signal volume due to the gain of the convertor.

Checking for signal presence, strength and quality at point 3 is more or less the same as checking point 2 except that we would use the audio input and probe of the signal tracer. Again, we are looking for a clear signal at the secondary of the second IF transformer, which is the same point, electrically, as the anode of the detector diode part of the 12AT6 (or more commonly, a 12AV6) tube. Signal failure or degradation here requires us to backtrack to the 12BA6, IF amplifier plate connection  to the primary of the second IF transformer to check the transformer. There we would use the tracer's RF probe and input again because the primary side of the transformer is still carrying unrectified, modulated IF frequency.  Of course this is just the way I prefer to do it, but one can also just go to the primaries/plates first, then the secondary sides, etc., and work toward the speaker, but I like to check the low voltage grid circuits and go back to check the high voltage plate circuits only if necessary. The signal after the IF amplifier, in any case, should be very much louder than before the IF stage. Note that the signal at the anode of the detector will usually tend to sound a bit "tinny" or trebly on a signal tracer in most receivers because of the tracer loading on the IF secondary, but should not be too rough or distorted. Again, this signal should be many times louder than the input of the IF amplifier. A signal tracer is a relative measurer of signal strength, and as imprecise as it may be, Figure 5 below is a chart that may help give an idea of relative signal loudness in a normally operating receiver, and this is assuming the input signal is unchanged and the volume control is turned full on. This is only an impression by my ears, but may be able to give some idea of what to expect if the receiver is working normally.

Figure 5 Relative Audibility Scale

---

Remember, test point 3 and after require the use of the audio signal function of the signal tracer because the IF signal has been rectified (demodulated) into audio at point 3. The volume control of the receiver must also be turned up to monitor signals at point 4 and after.
 
Point 4 is the grid of the 50C5 audio power amplifier tube. there should be a loud and good quality audio signal here. If not, there may be a problem with the coupling between the 12AT6,  1st audio tube and the power amplifier. If this is the case, we can go back to the input side of the coupling network to check the plate, and if necessary, the grid of the 1st audio amplifier tube. The signal level on the plate of the 1st audio tube should be much louder than at its grid, and the signal level on each side of the coupling network to the audio power tube should be approximately the same level with perhaps just a little loss, but generally an improvement in tone (less tinny) on the output side of the coupling network due to some frequency compensation there.  

Point 6 is the speaker terminals, and there will be a big drop in signal volume there because of the step down transformer used to match the output tube plate to the low impedance speaker windings. Point 7, will usually be the loudest signal in the receiver and will probably require you to turn down the volume control of the receiver, and perhaps the gain setting of the tracer because it can be annoyingly loud and often overdrive the tracer into distortion.

As always, when a signal is lost, partially or entirely,  or degraded in quality, the section before that test point is usually at fault. Start by checking voltages around the tubes. Tubes may be tested, or better yet, substituted with known good ones. Capacitors that may be open can be shunted with good capacitors of the same rating. Resistors values may be checked with the power off and disconnected from the line, but remember that any parallel path around a resistor will cause the resistance to read low, and many times it is necessary to clip one side of a resistor (or any component) loose in order to get a measurement. In old electronic equipment, the most typical failures are electrolytic type power supply capacitors drying up and loosing capacity to filter, or severely leaking current across them. Either can cause hum in the audio or an overload condition. (Receivers or amplifiers that use power transformers are vulnerable to transformer burnout from leaky or shorted power supply capacitors or a shorted rectifier tube.) Tube failures are not uncommon, of course, coupling and bypass capacitors—especially wax and paper type—leaking current, and resistors drifting out of spec—usually to a higher resistance. Any of these things can cause signal loss or degradation. Often, a wax and paper capacitor is used to block the plate DC from the grid of the following tube. These will tend to leak DC current or short entirely with age. A coupling capacitor that leaks even a tiny bit of DC current will tend to pull the grid of the following tube to a positive voltage, destroying its proper bias point and causing signal distortion on that following tube or even causing the tube to go into complete saturation and block the signal entirely. Power tubes, especially, can be overheated and destroyed in this way. Leaking or open bypass capacitors can cause similar conditions or squealing, howling, or "motorboating" due to positive feedback to the input of a previous circuit. The key to signal tracing or signal substitution is to try to isolate the likely problem area to a certain stage so that you are only dealing with a few components and voltages to check.

Signal substitution is another way of tracing a signal, but only a signal generator is required. The generator must have the capability to produce a variable strength radio frequency of the proper band that can be either amplitude modulated or unmodulated by the user. It will also need to be able to put out an audio frequency voltage of variable strength. Most signal generators for service work have a fixed audio output frequency of either 400 Hz or 1 kHz and this works fine. Since a signal generator of this type is practically a necessity for radio work such as doing alignments, most people seem to use the signal substitution (injection) method of troubleshooting. It seems less technicians and hobbyists own a signal tracer.

Troubleshooting by signal substitution (injection of a signal at various test points) starts at the speaker end of a receiver or amplifier because we are now using the device's own output as the indicator of the signal being passed along or not, or degraded or not. Naturally, we need to know that the speaker is good, first of all, so that's why we start at that end. Like signal tracing, the results will be relative, and it is not a bad idea to get some hands-on practice with a known good receiver or amplifier of similar type in order to know what is a normal response. Most signal generators put out a fairly feeble audio signal, and not all of them can drive a speaker directly. Typically though, putting an audio tone across the speaker terminals with the power removed from the device under test will yield a faintly audible tone. It depends on your signal generator. Continuing on, we will be using the AM radio receiver that was used in the previous examples about signal tracing. Testing the speaker with the power removed from the radio, we will assume that the generator produces a faint tone in the speaker when the audio output is placed directly across the speaker terminals. This indicates that the speaker is probably good, or at least can be used for finding where our signal is lost or degraded. Still with no power to the receiver, we would now move the audio signal to the primary side of the output transformer. In most cases this will produce a substantially louder tone from the speaker—or an audible tone if the generator failed to put out enough low impedance power to move the speaker cone—assuming the speaker and transformer are OK. If there is no tone here or directly at the speaker, substitute the generator with an ohmmeter or a 1.5V battery and make quick, temporary  contacts or "taps" with the same radio terminals to listen for clicks. We did not want or need power to the receiver to test the speaker and output transformer because the generator (or battery or ohmmeter)  is supplying all the audio power, weak as it may be, at these points. Before moving on, it is important to mention that not all signal generators provide a DC blocking capacitor in the audio (or even the RF) output terminal(s), and can therefore short circuit the device under test's power supply through the generator's output attenuator network. If in doubt use a capacitor in series with the "hot" lead of the generator. A 0.1 to 0.5 microfarad, 630V capacitor will work for most applications if its voltage rating is not exceeded be the device under test's power supply voltage.

Now moving backward to point 5, we would power up the receiver, turn up the volume control to avoid it shunting our signal to ground, and inject the generator's audio tone here. This should be quite substantially amplified be the output tube. If it sounds pretty rough, try turning down the test signal. If you cannot get a good volume level without distortion, and we are still assuming the generator has a blocking capacitor so it is not shorting the B+, you may suspect a defective power amp stage.
The sound level and performance at point 4 should be about the same since there is no amplification going on here. There will be some attenuation of the signal in the coupling network, and it will may have a little more bass (difficult to tell with a single tone, I know) because of the tone compensation network that is in the network to make the small speaker sound more natural and less "tinny."
Point 3 in this receiver will normally contain both audio and RF signal components when it is operating normally, and either an audio or modulated IF signal injected at this point will pass through the audio stages, but it is best to inject the modulated IF signal here. This is 455kHz in this and most AM broadcast band receivers, but there are sometimes others. There should be quite a lot of amplification from here—as long as you remembered to turn the volume control up before. Another point or two to test (not numbered) is the "top" of the volume control (the tube side of the variable resistor) and the wiper arm (center terminal) of the control. For these two we go back to the audio tone because the IF filters will take out most all of the high frequency here by shunting them to ground through small capacitors. Only audio should pass well and strong through most volume control circuits. Again, the 1st audio tube should provide a lot of signal gain over what level we had at points 4 and 5.

Point 2 should have a modulated 455kHz (or whatever the IF frequency) injected at a low level, because the IF amplifier provides the most overall signal gain of all the stages. If direct contact is made to the circuit (as opposed to just hovering the injection probe tip near the point 2) you may need to adjust the generator off the IF frequency a bit to get maximum receiver response. The capacitance of the test equipment will pull the tuning of the IF transformer off its 455kHz. Hovering the tip near the test point without contacting it directly will reduce this effect to almost nil. The same applies to the next test points or any tuned circuit points, but sometimes you have to make contact to get enough signal into the circuit under test.

The plate of the convertor tube (not numbered) should be similar to point 2, but there will be some attenuation of the signal across the IF transformer. This point will also need a modulated 455kHz signal applied to it.
Point 1 is a bit interesting because it should be responsive to the RF broadcast band frequencies of 550kHz to at least 1600kHz, but will also pass an IF signal, since this IF signal is normally generated in the tube by mixing the RF and oscillator outputs together. There should be a fair bit of gain in the convertor tube's output as well. For checking the frequency response (accuracy) of the tuning dial setting, we would need to inject the signal at a low level to the antenna through a few loops of wire in its proximity, or else the tuning will appear to be off frequency because of the signal generator's capacitive loading effect, as described before.
Well, that's all the main test point I have in this simple example radio receiver, and I will not go into any more detail regarding individual component or voltage analysis within a defective stage in this article. I hope it is sufficient, in spite of my ramblings and wanderings to give a baseline of information to build on. It is far better to systematically isolate a suspect area of a device first, than to look at the whole system as a big mess of wires and parts and scratch your head wondering what could be the problem. I hope this article from a moderately experienced fellow amateur helps to do that somewhat. Above all, know your electrical safety principles and procedures well, and use them!

© 2016 Rob's Radio-Active, LLC
All Rights Reserved